PROJECT 2 - PATTERN AND CELL SIGNALING IN MULTICELLULAR ORGANISMS Project Leader: E. Wieschaus (Molbio/LSI) During the development of multicellular organisms, cells are directed to specific fates and behaviors In response to signals in their environment, as well as signals they generate themselves. Although such signals produce patterns of extraordinary reproducibility and robustness, they require cells to measure subtle differences in concentration and to integrate those measurements with information from other signaling pathways and from the cell's own history. The process converts graded information into discrete changes in gene expression, cell shape and cellular behavior. We are interested in how signaling circuits are engineered to ensure accurate detection of signals, how cells integrate multiple signals into single discrete responses, and whether organisms tighten and relax their responses to signaling circuits to reflect the competing requirements for uniformity vs. diversity at different developmental stages. To approach these questions we have developed techniques to measure signaling cues and the corresponding responses in different organisms and at different stages of their development. Our experiments revisit well-studied genetically defined phenomena (aggregation in Dictyostelium, patterning in Drosophila ovaries and embryos, longevity in C. elegans) with the explicit goal of measuring both the numbers of molecules, the stochastic variability, and reproducibility of outcome in the context of mathematically testable models. The approach requires a broad range of expertise, is inherently interdisciplinary and depends on interactions fostered in the Center between physicists, chemical engineers and molecular biologists. One challenge faced by all signaling systems is that subtle changes must be measured in a background of random variation and noise. The low concentrations and corresponding small numbers of molecules involved often put severe limits on the ability to rapidly detect signals relative to noise. In slime molds, the transition from amoeboid to communal slug behaviors depends on a local response to waves of cAMP that polarize individual amoeba and reorganize the cytoskeleton at each cell's migratory leading edge. The initiating signals for this cAMP relay arise In oscillating centers that we have characterized in high through-put genetic screens following wave behaviors in individual colonies. Using a FRET based sensor to measure cytosolic cAMP, we have identified distinct oscillatory modes and characterized the rich dynamics that exists at the level of Individual cells. We have shown however that the absolute concentration of cAMP required for an individual cell's response is surprisingly low, suggesting that signaling in this system operate with a precision limited by the counting of individual molecules. Early development in Drosophila provides a second opportunity to study the precision of simple input/output signals. Here, we have studied the response of embryonic cells to a gradient of the maternal transcription factor Bicoid, using a fully functional EGFP tagged Bed protein to follow the dynamics of gradient formation and the response in individual cells along the anterior-posterior access of the egg. We found that the absolute concentration of Bed needed to activate one known target (Hunchback) is in the nanoMolar range, but that despite this low concentration, the Bed gradient is strikingly reproducible from embryo to embryo and the response of target genes in turn is so precise that concentration differences of just 10% are signaled reliably. These experiments led us to revisit the relevant theory of noise, identifying fundamental physical limits to precision in a series of progressively more complex and realistic models 185 Botstein, David The Drosophila embryo provides an opportunity to study integration of multiple simultaneous signals. Much of early zygotic gene expression depends on inputs from multiple maternal sources and once target genes are expressed, their product may Interact In complex regulatory circuits at the gap gene and pair-rule gene levels. We do not know whether the same tight input/output relationships are maintained at these levels and under conditions when responses depend on multiple inputs. We outline below experiments to address these issues. The ovarian follicle cell epithelium in Drosophila provides a second (and perhaps simpler system) of signal integration because spatial pattern is determined by two discrete signaling systems that define anterior posterior and dorsal ventral position. Using large-scale microarray experiments, we have identified the EGF and Dpp targets that function in this pathway and the cross talk that regulates their expression. These genes are expressed in patterns that could be described using a compact combinatorial code that is based on four algebraic operations and only six geometric shapes (primitives) derived from the distribution of the patterning signals. In most of our input/output analysis, we have used immediate biochemical changes in second messengers, or changes in transcription as the readout assays. These immediate responses are rapidly converted into discrete cellular behaviors, and it is the robustness of those behaviors that is functionally significant and thus the target of evolutionary selection. Such cellular behaviors are often more complex, involve multiple redundant components and thus require integration at a higher level than immediate responses. Such cellular responses can be studied in slime molds and in Drosophila blastoderm, and we propose to develop the follicle cell epithelium as an additional system for studying integrated changes in cell morphology in response to signaling. Morphological change depends on local force generation and the most immediate challenge will be to connect the discrete expression-based signal responses to the cell biological machines that generate graded differences in force. This will require both accurate measurement of cell behavior and mechanical models to test hypothetical connections between graded cell signals, discrete responses in terms of local patterns of gene expression, and mechanical properties that may be again graded or localized. During development graded signals are converted to robust cellular responses that are discrete and reproducible. Precise Invariant behaviors may serve the developing organism well at certain stages, but may not always be essential or even useful. Longevity in C. elegans may provide an interesting opportunity to study how tightly controlled cell lineage and developmental cell tjehavlors can be relaxed to allow stochastic variability between individuals with a uniform population. One advantage of C. elegans for these studies is that the same organism can show extremely reproducible behaviors at one stage In development and increased variability at another. To address these questions we have developed tools and new technologies that will be broadly useful to the scientific community as a whole. This development was only possible in the context of the shared facilities and intellectual interactions supported by the Center. In our work on Dictyostelium, we developed a robotic microscope for recording the aggregation and life cycle behavior of large numbers of individual colonies and plan in the next granting period to adapt that robot for analysis of life cycle behaviors in individual C. elegans. Our past analysis of Bed distribution was carried out using a specially adapted two photon microscope build in the Center imaging facility. For the more detailed analysis of 3D gene expression patterns described in this proposal, we will need to build a special microscope in which Drosophila samples can be rotated 360 for data collection.